{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "b3b01615",
   "metadata": {},
   "outputs": [],
   "source": [
    "#### In this script, we run a robustness test based on the thresholds of the previous Fields of Study (FOS) of the paper authors.\n",
    "#### The idea being, if we used different thresholds to determine interdisciplinary categories, would we get the same results?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "42dd44d3",
   "metadata": {},
   "outputs": [],
   "source": [
    "### ALL IMPORTS\n",
    "\n",
    "import os\n",
    "import pandas as pd\n",
    "import re\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from PIL import Image\n",
    "\n",
    "from statsmodels.api import OLS, add_constant\n",
    "from scipy import stats\n",
    "import numpy as np\n",
    "from statsmodels.stats.diagnostic import het_breuschpagan\n",
    "from statsmodels.stats.outliers_influence import variance_inflation_factor\n",
    "from statsmodels.stats.stattools import durbin_watson\n",
    "from scipy.stats import linregress\n",
    "from linearmodels.panel import PanelOLS\n",
    "from linearmodels.panel import compare"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b590c412",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(99501, 29)\n"
     ]
    }
   ],
   "source": [
    "### Load the aggregated data to link to the paper-level data with the FOS thresholds\n",
    "\n",
    "### directories\n",
    "\n",
    "DIR = \"/Replication_SocietalAI\"\n",
    "DATA_DIR = os.path.join(DIR, \"Data\")\n",
    "GRAPHICS_DIR = os.path.join(DIR, \"Graphics\")\n",
    "if not os.path.exists(GRAPHICS_DIR):\n",
    "    os.makedirs(GRAPHICS_DIR)\n",
    "##### Load the main file\n",
    "df = pd.read_csv(os.path.join(DATA_DIR, \"processed_June2025_final_data.csv\"))\n",
    "print(df.shape)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "a41f0b1d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['Unnamed: 0.1', 'Unnamed: 0', 'abstract', 'author_ids',\n",
       "       'citation_count', 'conclusion', 'introduction', 'authors_final_FOS',\n",
       "       'paper_id', 'percent_societal', 'RQ_societal', 'RQ', 'subdomain',\n",
       "       'title', 'venue', 'year', 'team_size', 'event_type',\n",
       "       'percent_societal_new', 'total_sentences', 'article_length',\n",
       "       'categories', 'cs.cv', 'cs.cl', 'cs.cy', 'cs.ai', 'extracted_FOS',\n",
       "       'standardized_FOS', 'interdisciplinary_category'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "87614035",
   "metadata": {},
   "outputs": [],
   "source": [
    "### prepare the meta columns for the regression later on\n",
    "\n",
    "# Split the categories column into individual categories\n",
    "categories_split = df['categories'].str.split()\n",
    "\n",
    "# Filter rows to only include articles with the desired categories\n",
    "categories_of_interest = {'cs.cl', 'cs.cv', 'cs.cy', 'cs.ai'}\n",
    "filtered_df = df[\n",
    "    categories_split.apply(lambda x: any(category in categories_of_interest for category in x))\n",
    "]\n",
    "\n",
    "# Create dummy variables for each unique category\n",
    "dummies = pd.get_dummies(categories_split.apply(pd.Series).stack()).groupby(level=0).sum()\n",
    "\n",
    "# Filter dummy variables to include only the categories of interest\n",
    "dummies = dummies[list(categories_of_interest)]\n",
    "\n",
    "# Merge the filtered dummy variables back into the filtered DataFrame\n",
    "filtered_df = pd.concat([filtered_df, dummies], axis=1)\n",
    "\n",
    "df = filtered_df.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "190da481",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['FOS_75.csv', '.DS_Store', 'FOS_80.csv']\n"
     ]
    }
   ],
   "source": [
    "#### load the other threshold files:\n",
    "\n",
    "THRESH_DIR = os.path.join(DATA_DIR, \"Robustness/FOS\")\n",
    "print(os.listdir(THRESH_DIR))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "710149fb",
   "metadata": {},
   "outputs": [],
   "source": [
    "#### load the new author info based on different thresholds\n",
    "\n",
    "df_75 = pd.read_csv(os.path.join(THRESH_DIR, \"FOS_75.csv\"))\n",
    "df_80 = pd.read_csv(os.path.join(THRESH_DIR, \"FOS_80.csv\"))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "ea2ce328",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(99428, 34)\n",
      "(99428, 35)\n"
     ]
    }
   ],
   "source": [
    "### inner merge data with df_75 and df_80 on \"paper_id\"\n",
    "merged = df.merge(df_75[['paper_id','threshold_75']], on='paper_id', how='inner')\n",
    "print(merged.shape)\n",
    "merged = merged.merge(df_80[['paper_id','threshold_80']], on='paper_id', how='inner')\n",
    "print(merged.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "8c5e2a42",
   "metadata": {},
   "outputs": [],
   "source": [
    "#### change back variable name for clarity now that merged all data\n",
    "df = merged.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "e4fbf85c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Data shape after interdisciplinary classification: (99428, 41)\n",
      "Data shape after removing unclassified rows: (99428, 41)\n",
      "Data shape after removing unclassified rows: (99309, 41)\n",
      "Data shape after removing unclassified rows: (99308, 41)\n",
      "Data interdisicplinary category counts:\n",
      "interdisciplinary_category\n",
      "CS-only                                           74639\n",
      "Interdisciplinary (Natural Sciences/Medicine)     12354\n",
      "Fully Interdisciplinary                            7853\n",
      "Interdisciplinary (Social Sciences/Humanities)     4462\n",
      "Name: count, dtype: int64\n",
      "Data 75 interdisicplinary category counts:\n",
      "interdisciplinary_category_75\n",
      "CS-only                                           71163\n",
      "Interdisciplinary (Natural Sciences/Medicine)     10987\n",
      "Interdisciplinary (Social Sciences/Humanities)     8779\n",
      "Fully Interdisciplinary                            8379\n",
      "Name: count, dtype: int64\n",
      "Data 80 interdisicplinary category counts:\n",
      "interdisciplinary_category_80\n",
      "CS-only                                           71121\n",
      "Interdisciplinary (Natural Sciences/Medicine)     11043\n",
      "Interdisciplinary (Social Sciences/Humanities)     8774\n",
      "Fully Interdisciplinary                            8370\n",
      "Name: count, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# ---------------------------\n",
    "# Define Interdisciplinary Categories\n",
    "# ---------------------------\n",
    "\n",
    "# Define disciplinary fields by area\n",
    "cs_fields = {'Computer Science', 'Engineering', 'Mathematics', 'Physics', 'Materials Science'}\n",
    "social_fields = {'Political Science', 'Sociology', 'Philosophy', 'History', 'Business', 'Art', 'Economics', 'Geography'}\n",
    "natural_fields = {'Medicine', 'Biology', 'Chemistry', 'Environmental Science', 'Geology', 'Psychology'}\n",
    "\n",
    "# Standardize and classify Fields of Study\n",
    "core_fields = cs_fields.union(social_fields).union(natural_fields)\n",
    "pattern = re.compile(r'\\b(' + '|'.join(core_fields) + r')\\b', re.IGNORECASE)\n",
    "\n",
    "# Extract fields from the main_FOS column\n",
    "def extract_fos(text):\n",
    "    matches = pattern.findall(str(text))\n",
    "    return ', '.join(matches) if matches else 'Unknown'\n",
    "\n",
    "def standardize_fos(text):\n",
    "    if text == 'Unknown':\n",
    "        return 'Unknown'\n",
    "    parts = [p.strip() for p in re.split(r',\\s*', str(text)) if p.strip()]\n",
    "    return ', '.join(sorted(set(parts)))\n",
    "\n",
    "\n",
    "# Apply extraction and standardization to data\n",
    "df['extracted_FOS_75'] = df['threshold_75'].apply(extract_fos)\n",
    "df['standardized_FOS_75'] = df['extracted_FOS_75'].apply(standardize_fos)\n",
    "\n",
    "# Apply extraction and standardization to data\n",
    "df['extracted_FOS_80'] = df['threshold_80'].apply(extract_fos)\n",
    "df['standardized_FOS_80'] = df['extracted_FOS_80'].apply(standardize_fos)\n",
    "\n",
    "# Classify interdisciplinary team composition\n",
    "def classify_interdisciplinarity(fos_list):\n",
    "    fos_set = set(fos_list.split(', '))\n",
    "    cs_flag = any(f in cs_fields for f in fos_set)\n",
    "    social_flag = any(f in social_fields for f in fos_set)\n",
    "    natural_flag = any(f in natural_fields for f in fos_set)\n",
    "\n",
    "    if cs_flag and social_flag and natural_flag:\n",
    "        return 'Fully Interdisciplinary'\n",
    "    elif social_flag and not natural_flag:\n",
    "        return 'Interdisciplinary (Social Sciences/Humanities)'\n",
    "    elif natural_flag and not social_flag:\n",
    "        return 'Interdisciplinary (Natural Sciences/Medicine)'\n",
    "    elif cs_flag:\n",
    "        return 'CS-only'\n",
    "    return 'Unclassified'\n",
    "\n",
    "# Add classification to the dataframe\n",
    "df['interdisciplinary_category'] = df['standardized_FOS'].apply(classify_interdisciplinarity)\n",
    "df['interdisciplinary_category_75'] = df['standardized_FOS_75'].apply(classify_interdisciplinarity)\n",
    "df['interdisciplinary_category_80'] = df['standardized_FOS_80'].apply(classify_interdisciplinarity)\n",
    "\n",
    "print(f\"Data shape after interdisciplinary classification: {df.shape}\")\n",
    "# Remove unclassified rows for cleaner regression input\n",
    "df = df[df['interdisciplinary_category'] != 'Unclassified']\n",
    "print(f\"Data shape after removing unclassified rows: {df.shape}\")\n",
    "\n",
    "# Remove unclassified rows for cleaner regression input\n",
    "df = df[df['interdisciplinary_category_75'] != 'Unclassified']\n",
    "print(f\"Data shape after removing unclassified rows: {df.shape}\")\n",
    "\n",
    "df = df[df['interdisciplinary_category_80'] != 'Unclassified']\n",
    "print(f\"Data shape after removing unclassified rows: {df.shape}\")\n",
    "\n",
    "print('Data interdisicplinary category counts:')\n",
    "print(df['interdisciplinary_category'].value_counts())\n",
    "\n",
    "\n",
    "print('Data 75 interdisicplinary category counts:')\n",
    "print(df['interdisciplinary_category_75'].value_counts())\n",
    "\n",
    "print('Data 80 interdisicplinary category counts:')\n",
    "print(df['interdisciplinary_category_80'].value_counts())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "90cdbb52",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "=== Averages for Original ===\n",
      "     Team Type  Avg Sentence-Level (%)  Avg RQ-Level (%)\n",
      "       CS-only                    7.77              0.08\n",
      "Interd. (Full)                   12.42              0.26\n",
      " Interd. (NSM)                   13.13              0.19\n",
      " Interd. (SSH)                   20.72              0.35\n",
      "\n",
      "=== Averages for 75 Threshold ===\n",
      "     Team Type  Avg Sentence-Level (%)  Avg RQ-Level (%)\n",
      "       CS-only                    7.45              0.08\n",
      "Interd. (Full)                   15.94              0.30\n",
      " Interd. (NSM)                   10.04              0.15\n",
      " Interd. (SSH)                   17.99              0.27\n",
      "\n",
      "=== Averages for 80 Threshold ===\n",
      "     Team Type  Avg Sentence-Level (%)  Avg RQ-Level (%)\n",
      "       CS-only                    7.45              0.08\n",
      "Interd. (Full)                   15.91              0.30\n",
      " Interd. (NSM)                   10.02              0.15\n",
      " Interd. (SSH)                   17.99              0.27\n"
     ]
    }
   ],
   "source": [
    "# Map for display\n",
    "label_map_display = {\n",
    "    'CS-only': 'CS-only',\n",
    "    'Interdisciplinary (Social Sciences/Humanities)': 'Interd. (SSH)',\n",
    "    'Interdisciplinary (Natural Sciences/Medicine)': 'Interd. (NSM)',\n",
    "    'Fully Interdisciplinary': 'Interd. (Full)'\n",
    "}\n",
    "\n",
    "# Function to prepare and summarize one threshold\n",
    "def summarize_threshold(df, category_column, threshold_name):\n",
    "    temp = df.copy()\n",
    "    temp['RQ_societal'] = temp['RQ_societal'].map({'Yes': 1, 'No': 0})\n",
    "    temp = temp.dropna(subset=['RQ_societal'])\n",
    "    temp['RQ_societal'] = temp['RQ_societal'].astype(int)\n",
    "    temp['short_label'] = temp[category_column].map(label_map_display)\n",
    "    temp = temp[temp['short_label'].notnull()]\n",
    "\n",
    "    means = temp.groupby('short_label')[['percent_societal', 'RQ_societal']].mean().reset_index()\n",
    "    means.columns = ['Team Type', 'Avg Sentence-Level (%)', 'Avg RQ-Level (%)']\n",
    "    print(f\"\\n=== Averages for {threshold_name} ===\")\n",
    "    print(means.to_string(index=False, float_format='%.2f'))\n",
    "\n",
    "# Original\n",
    "summarize_threshold(df, 'interdisciplinary_category', 'Original')\n",
    "\n",
    "# 75 Threshold\n",
    "summarize_threshold(df, 'interdisciplinary_category_75', '75 Threshold')\n",
    "\n",
    "# 80 Threshold\n",
    "summarize_threshold(df, 'interdisciplinary_category_80', '80 Threshold')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "cb731bd0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x1000 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Font constants\n",
    "TITLE_SIZE = 24\n",
    "LABEL_SIZE = 21\n",
    "TICK_SIZE = 16\n",
    "LEGEND_SIZE = 15\n",
    "LEGEND_TITLE_SIZE = 14\n",
    "\n",
    "# Label mapping\n",
    "label_map_display = {\n",
    "    'CS-only': 'CS-only',\n",
    "    'Interdisciplinary (Social Sciences/Humanities)': 'Interd. (SSH)',\n",
    "    'Interdisciplinary (Natural Sciences/Medicine)': 'Interd. (NSM)',\n",
    "    'Fully Interdisciplinary': 'Interd. (Full)'\n",
    "}\n",
    "\n",
    "# Function to prepare one threshold\n",
    "def prep_df(df, category_column, threshold_name):\n",
    "    temp = df.copy()\n",
    "    temp['RQ_societal'] = temp['RQ_societal'].map({'Yes': 1, 'No': 0})\n",
    "    temp = temp.dropna(subset=['RQ_societal'])\n",
    "    temp['RQ_societal'] = temp['RQ_societal'].astype(int)\n",
    "    temp['short_label'] = temp[category_column].map(label_map_display)\n",
    "    temp = temp[temp['short_label'].notnull()]\n",
    "    temp['threshold'] = threshold_name\n",
    "    return temp[['short_label', 'percent_societal', 'RQ_societal', 'threshold']]\n",
    "\n",
    "# Prepare data for each threshold\n",
    "df_orig = prep_df(df, 'interdisciplinary_category', 'Original')\n",
    "df_75 = prep_df(df, 'interdisciplinary_category_75', '75 Threshold')\n",
    "df_80 = prep_df(df, 'interdisciplinary_category_80', '80 Threshold')\n",
    "\n",
    "# Combine\n",
    "combined_df = pd.concat([df_orig, df_75, df_80], axis=0)\n",
    "\n",
    "# Define palette\n",
    "cb_palette = {\n",
    "    'Original': '#0072B2',       # dark blue\n",
    "    '80 Threshold': '#E69F00',   # orange\n",
    "    '75 Threshold': '#009E73'    # green\n",
    "}\n",
    "\n",
    "# Plot setup\n",
    "sns.set(style='whitegrid')\n",
    "fig, axes = plt.subplots(2, 1, figsize=(10, 10), sharex=True)\n",
    "\n",
    "# Document-level plot\n",
    "sns.barplot(\n",
    "    data=combined_df,\n",
    "    x='short_label',\n",
    "    y='percent_societal',\n",
    "    hue='threshold',\n",
    "    ax=axes[0],\n",
    "    palette=cb_palette,\n",
    "    errorbar=('ci', 95)\n",
    ")\n",
    "axes[0].set_title('A. Document-Level Societal Orientation', fontsize=TITLE_SIZE)\n",
    "axes[0].set_ylabel('Avg Percent', fontsize=LABEL_SIZE)\n",
    "axes[0].tick_params(axis='both', labelsize=TICK_SIZE)\n",
    "axes[0].legend(title='Threshold', fontsize=LEGEND_SIZE, title_fontsize=LEGEND_TITLE_SIZE)\n",
    "\n",
    "# RQ-level plot\n",
    "sns.barplot(\n",
    "    data=combined_df,\n",
    "    x='short_label',\n",
    "    y='RQ_societal',\n",
    "    hue='threshold',\n",
    "    ax=axes[1],\n",
    "    palette=cb_palette,\n",
    "    errorbar=('ci', 95)\n",
    ")\n",
    "axes[1].set_title('B. Societal Research Questions', fontsize=TITLE_SIZE)\n",
    "axes[1].set_ylabel('Percent', fontsize=LABEL_SIZE)\n",
    "axes[1].set_xlabel('', fontsize=LABEL_SIZE)\n",
    "axes[1].tick_params(axis='both', labelsize=TICK_SIZE)\n",
    "axes[1].legend(title='Threshold', fontsize=LEGEND_SIZE, title_fontsize=LEGEND_TITLE_SIZE)\n",
    "plt.setp(axes[1].get_xticklabels(), rotation=20)\n",
    "\n",
    "# Save figures\n",
    "plt.tight_layout()\n",
    "color_path = f\"{GRAPHICS_DIR}/avg_societal_orientation_thresholds.png\"\n",
    "gray_path = f\"{GRAPHICS_DIR}/avg_societal_orientation_thresholds_grayscale.png\"\n",
    "\n",
    "plt.savefig(f\"{GRAPHICS_DIR}/avg_societal_orientation_thresholds.pdf\", format='pdf', dpi=300)\n",
    "plt.savefig(color_path, format='png', dpi=300, bbox_inches='tight')\n",
    "\n",
    "# Convert to grayscale\n",
    "img = Image.open(color_path).convert(\"L\")\n",
    "img.save(gray_path)\n",
    "\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "ff6c4065",
   "metadata": {},
   "outputs": [],
   "source": [
    "########################################\n",
    "######## REGRESSION ANALYSIS ###########\n",
    "########################################"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "1139062f",
   "metadata": {},
   "outputs": [],
   "source": [
    "### drop duplicated columns\n",
    "df = df.loc[:, ~df.columns.duplicated()]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "1e590ea6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "PanelOLS Results with Robust Standard Errors:\n",
      "                          PanelOLS Estimation Summary                           \n",
      "================================================================================\n",
      "Dep. Variable:       percent_societal   R-squared:                        0.3731\n",
      "Estimator:                   PanelOLS   R-squared (Between):              0.1246\n",
      "No. Observations:               99308   R-squared (Within):               0.3654\n",
      "Date:                Thu, Jul 03 2025   R-squared (Overall):              0.3445\n",
      "Time:                        16:47:42   Log-likelihood                -3.949e+05\n",
      "Cov. Estimator:                Robust                                           \n",
      "                                        F-statistic:                      6564.5\n",
      "Entities:                           2   P-value                           0.0000\n",
      "Avg Obs:                    4.965e+04   Distribution:                 F(9,99287)\n",
      "Min Obs:                       1919.0                                           \n",
      "Max Obs:                    9.739e+04   F-statistic (robust):             2301.3\n",
      "                                        P-value                           0.0000\n",
      "Time periods:                      11   Distribution:                 F(9,99287)\n",
      "Avg Obs:                       9028.0                                           \n",
      "Min Obs:                       2135.0                                           \n",
      "Max Obs:                    1.988e+04                                           \n",
      "                                                                                \n",
      "                                               Parameter Estimates                                                \n",
      "==================================================================================================================\n",
      "                                                Parameter  Std. Err.     T-stat    P-value    Lower CI    Upper CI\n",
      "------------------------------------------------------------------------------------------------------------------\n",
      "Fully Interdisciplinary                            4.9953     0.2171     23.004     0.0000      4.5697      5.4209\n",
      "Interdisciplinary (Natural Sciences/Medicine)      3.3250     0.1619     20.537     0.0000      3.0077      3.6423\n",
      "Interdisciplinary (Social Sciences/Humanities)     5.5658     0.2948     18.879     0.0000      4.9880      6.1436\n",
      "team_size                                          0.1240     0.0265     4.6786     0.0000      0.0720      0.1759\n",
      "article_length                                    -0.0003  2.223e-05    -14.581     0.0000     -0.0004     -0.0003\n",
      "cs.cl                                             -3.8191     0.1869    -20.438     0.0000     -4.1854     -3.4529\n",
      "cs.cv                                             -3.4605     0.1519    -22.781     0.0000     -3.7582     -3.1628\n",
      "cs.ai                                              2.0220     0.1259     16.060     0.0000      1.7753      2.2688\n",
      "cs.cy                                              31.197     0.2941     106.06     0.0000      30.621      31.774\n",
      "==================================================================================================================\n",
      "\n",
      "F-test for Poolability: 148.86\n",
      "P-value: 0.0000\n",
      "Distribution: F(11,99287)\n",
      "\n",
      "Included effects: Entity, Time\n",
      "\n",
      "PanelOLS Results with Clustered Standard Errors (by event_type):\n",
      "                          PanelOLS Estimation Summary                           \n",
      "================================================================================\n",
      "Dep. Variable:       percent_societal   R-squared:                        0.3731\n",
      "Estimator:                   PanelOLS   R-squared (Between):              0.1246\n",
      "No. Observations:               99308   R-squared (Within):               0.3654\n",
      "Date:                Thu, Jul 03 2025   R-squared (Overall):              0.3445\n",
      "Time:                        16:47:42   Log-likelihood                -3.949e+05\n",
      "Cov. Estimator:             Clustered                                           \n",
      "                                        F-statistic:                      6564.5\n",
      "Entities:                           2   P-value                           0.0000\n",
      "Avg Obs:                    4.965e+04   Distribution:                 F(9,99287)\n",
      "Min Obs:                       1919.0                                           \n",
      "Max Obs:                    9.739e+04   F-statistic (robust):         -3.058e+19\n",
      "                                        P-value                           1.0000\n",
      "Time periods:                      11   Distribution:                 F(9,99287)\n",
      "Avg Obs:                       9028.0                                           \n",
      "Min Obs:                       2135.0                                           \n",
      "Max Obs:                    1.988e+04                                           \n",
      "                                                                                \n",
      "                                               Parameter Estimates                                                \n",
      "==================================================================================================================\n",
      "                                                Parameter  Std. Err.     T-stat    P-value    Lower CI    Upper CI\n",
      "------------------------------------------------------------------------------------------------------------------\n",
      "Fully Interdisciplinary                            4.9953     0.0982     50.854     0.0000      4.8028      5.1878\n",
      "Interdisciplinary (Natural Sciences/Medicine)      3.3250     0.0905     36.748     0.0000      3.1476      3.5023\n",
      "Interdisciplinary (Social Sciences/Humanities)     5.5658     0.1793     31.049     0.0000      5.2144      5.9171\n",
      "team_size                                          0.1240     0.0347     3.5703     0.0004      0.0559      0.1920\n",
      "article_length                                    -0.0003  1.734e-06    -186.88     0.0000     -0.0003     -0.0003\n",
      "cs.cl                                             -3.8191     0.0825    -46.273     0.0000     -3.9809     -3.6573\n",
      "cs.cv                                             -3.4605     0.0719    -48.110     0.0000     -3.6015     -3.3195\n",
      "cs.ai                                              2.0220     0.0243     83.239     0.0000      1.9744      2.0696\n",
      "cs.cy                                              31.197     0.1007     309.68     0.0000      31.000      31.394\n",
      "==================================================================================================================\n",
      "\n",
      "F-test for Poolability: 148.86\n",
      "P-value: 0.0000\n",
      "Distribution: F(11,99287)\n",
      "\n",
      "Included effects: Entity, Time\n"
     ]
    }
   ],
   "source": [
    "#### normal threshold\n",
    "\n",
    "### Main research questions - interdisciplinarity and societal engagement\n",
    "\n",
    "# Ensure year is an integer (PanelOLS requires numeric time index)\n",
    "df['year'] = df['year'].astype(int)\n",
    "\n",
    "# Set a panel index: 'event_type' is the entity effect, 'year' is the time effect\n",
    "data_panel = df.set_index(['event_type', 'year'])\n",
    "\n",
    "# Define dependent and independent variables\n",
    "Y = data_panel['percent_societal']\n",
    "X = pd.get_dummies(data_panel['interdisciplinary_category'], drop_first=True)  # Use CS-only as baseline\n",
    "X['team_size'] = data_panel['team_size']\n",
    "X['article_length'] = data_panel['article_length']\n",
    "selected_categories = {'cs.cl', 'cs.cv', 'cs.cy', 'cs.ai'}  # Define which subfields to track\n",
    "selected_categories = list(selected_categories)\n",
    "\n",
    "# Add AI subfield controls if present\n",
    "for category in selected_categories:\n",
    "    if category in data_panel.columns:\n",
    "        X[category] = data_panel[category]\n",
    "\n",
    "# ---------------------------\n",
    "# Estimate Models\n",
    "# ---------------------------\n",
    "\n",
    "# Define the base model with fixed effects\n",
    "panel_model = PanelOLS(Y, X, entity_effects=True, time_effects=True)\n",
    "\n",
    "# Model 1: Robust standard errors (default recommended)\n",
    "results_robust = panel_model.fit(cov_type='robust')\n",
    "print(\"\\nPanelOLS Results with Robust Standard Errors:\")\n",
    "print(results_robust.summary)\n",
    "\n",
    "# Model 2: Clustered standard errors (not reliable with only 2 clusters)\n",
    "clusters = pd.Series(data_panel.index.get_level_values('event_type'), index=data_panel.index)\n",
    "results_clustered = panel_model.fit(cov_type='clustered', clusters=clusters)\n",
    "print(\"\\nPanelOLS Results with Clustered Standard Errors (by event_type):\")\n",
    "print(results_clustered.summary)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "70377132",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "PanelOLS Results with Robust Standard Errors for 75 Threshold:\n",
      "                          PanelOLS Estimation Summary                           \n",
      "================================================================================\n",
      "Dep. Variable:       percent_societal   R-squared:                        0.3674\n",
      "Estimator:                   PanelOLS   R-squared (Between):              0.0488\n",
      "No. Observations:               99308   R-squared (Within):               0.3609\n",
      "Date:                Thu, Jul 03 2025   R-squared (Overall):              0.3231\n",
      "Time:                        16:47:42   Log-likelihood                -3.954e+05\n",
      "Cov. Estimator:                Robust                                           \n",
      "                                        F-statistic:                      6406.2\n",
      "Entities:                           2   P-value                           0.0000\n",
      "Avg Obs:                    4.965e+04   Distribution:                 F(9,99287)\n",
      "Min Obs:                       1919.0                                           \n",
      "Max Obs:                    9.739e+04   F-statistic (robust):             2200.7\n",
      "                                        P-value                           0.0000\n",
      "Time periods:                      11   Distribution:                 F(9,99287)\n",
      "Avg Obs:                       9028.0                                           \n",
      "Min Obs:                       2135.0                                           \n",
      "Max Obs:                    1.988e+04                                           \n",
      "                                                                                \n",
      "                                               Parameter Estimates                                                \n",
      "==================================================================================================================\n",
      "                                                Parameter  Std. Err.     T-stat    P-value    Lower CI    Upper CI\n",
      "------------------------------------------------------------------------------------------------------------------\n",
      "Fully Interdisciplinary                            2.2387     0.2167     10.330     0.0000      1.8140      2.6635\n",
      "Interdisciplinary (Natural Sciences/Medicine)      1.3485     0.1555     8.6730     0.0000      1.0438      1.6532\n",
      "Interdisciplinary (Social Sciences/Humanities)     2.1679     0.1917     11.308     0.0000      1.7922      2.5436\n",
      "team_size                                          0.1423     0.0307     4.6360     0.0000      0.0821      0.2024\n",
      "article_length                                    -0.0003  1.982e-05    -13.682     0.0000     -0.0003     -0.0002\n",
      "cs.cl                                             -2.7129     0.1814    -14.953     0.0000     -3.0686     -2.3573\n",
      "cs.cv                                             -3.8738     0.1506    -25.717     0.0000     -4.1690     -3.5786\n",
      "cs.ai                                              1.5227     0.1246     12.222     0.0000      1.2785      1.7669\n",
      "cs.cy                                              31.632     0.3005     105.27     0.0000      31.043      32.221\n",
      "==================================================================================================================\n",
      "\n",
      "F-test for Poolability: 131.85\n",
      "P-value: 0.0000\n",
      "Distribution: F(11,99287)\n",
      "\n",
      "Included effects: Entity, Time\n",
      "\n",
      "PanelOLS Results with Clustered Standard Errors (by event_type):\n",
      "                          PanelOLS Estimation Summary                           \n",
      "================================================================================\n",
      "Dep. Variable:       percent_societal   R-squared:                        0.3674\n",
      "Estimator:                   PanelOLS   R-squared (Between):              0.0488\n",
      "No. Observations:               99308   R-squared (Within):               0.3609\n",
      "Date:                Thu, Jul 03 2025   R-squared (Overall):              0.3231\n",
      "Time:                        16:47:42   Log-likelihood                -3.954e+05\n",
      "Cov. Estimator:             Clustered                                           \n",
      "                                        F-statistic:                      6406.2\n",
      "Entities:                           2   P-value                           0.0000\n",
      "Avg Obs:                    4.965e+04   Distribution:                 F(9,99287)\n",
      "Min Obs:                       1919.0                                           \n",
      "Max Obs:                    9.739e+04   F-statistic (robust):          2.118e+20\n",
      "                                        P-value                           0.0000\n",
      "Time periods:                      11   Distribution:                 F(9,99287)\n",
      "Avg Obs:                       9028.0                                           \n",
      "Min Obs:                       2135.0                                           \n",
      "Max Obs:                    1.988e+04                                           \n",
      "                                                                                \n",
      "                                               Parameter Estimates                                                \n",
      "==================================================================================================================\n",
      "                                                Parameter  Std. Err.     T-stat    P-value    Lower CI    Upper CI\n",
      "------------------------------------------------------------------------------------------------------------------\n",
      "Fully Interdisciplinary                            2.2387     0.0280     79.910     0.0000      2.1838      2.2936\n",
      "Interdisciplinary (Natural Sciences/Medicine)      1.3485     0.0749     18.000     0.0000      1.2017      1.4953\n",
      "Interdisciplinary (Social Sciences/Humanities)     2.1679     0.0688     31.501     0.0000      2.0330      2.3028\n",
      "team_size                                          0.1423     0.0411     3.4633     0.0005      0.0617      0.2228\n",
      "article_length                                    -0.0003  5.668e-07    -478.36     0.0000     -0.0003     -0.0003\n",
      "cs.cl                                             -2.7129     0.0437    -62.054     0.0000     -2.7986     -2.6273\n",
      "cs.cv                                             -3.8738     0.0831    -46.595     0.0000     -4.0368     -3.7109\n",
      "cs.ai                                              1.5227     0.0340     44.787     0.0000      1.4561      1.5894\n",
      "cs.cy                                              31.632     0.0727     435.07     0.0000      31.489      31.774\n",
      "==================================================================================================================\n",
      "\n",
      "F-test for Poolability: 131.85\n",
      "P-value: 0.0000\n",
      "Distribution: F(11,99287)\n",
      "\n",
      "Included effects: Entity, Time\n"
     ]
    }
   ],
   "source": [
    "#### 75% threshold\n",
    "\n",
    "### Main research questions - interdisciplinarity and societal engagement\n",
    "\n",
    "# Define dependent and independent variables\n",
    "Y = data_panel['percent_societal']\n",
    "X = pd.get_dummies(data_panel['interdisciplinary_category_75'], drop_first=True)  # Use CS-only as baseline\n",
    "X['team_size'] = data_panel['team_size']\n",
    "X['article_length'] = data_panel['article_length']\n",
    "\n",
    "# Add AI subfield controls if present\n",
    "for category in selected_categories:\n",
    "    if category in data_panel.columns:\n",
    "        X[category] = data_panel[category]\n",
    "\n",
    "# ---------------------------\n",
    "# Estimate Models\n",
    "# ---------------------------\n",
    "\n",
    "# Define the base model with fixed effects\n",
    "panel_model = PanelOLS(Y, X, entity_effects=True, time_effects=True)\n",
    "\n",
    "# Model 1: Robust standard errors (default recommended)\n",
    "results_robust = panel_model.fit(cov_type='robust')\n",
    "print(\"\\nPanelOLS Results with Robust Standard Errors for 75 Threshold:\")\n",
    "print(results_robust.summary)\n",
    "\n",
    "# Model 2: Clustered standard errors (not reliable with only 2 clusters)\n",
    "clusters = pd.Series(data_panel.index.get_level_values('event_type'), index=data_panel.index)\n",
    "results_clustered = panel_model.fit(cov_type='clustered', clusters=clusters)\n",
    "print(\"\\nPanelOLS Results with Clustered Standard Errors (by event_type):\")\n",
    "print(results_clustered.summary)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "542705b5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "PanelOLS Results with Robust Standard Errors for 80 Threshold:\n",
      "                          PanelOLS Estimation Summary                           \n",
      "================================================================================\n",
      "Dep. Variable:       percent_societal   R-squared:                        0.3674\n",
      "Estimator:                   PanelOLS   R-squared (Between):              0.0493\n",
      "No. Observations:               99308   R-squared (Within):               0.3609\n",
      "Date:                Thu, Jul 03 2025   R-squared (Overall):              0.3232\n",
      "Time:                        16:47:42   Log-likelihood                -3.954e+05\n",
      "Cov. Estimator:                Robust                                           \n",
      "                                        F-statistic:                      6406.4\n",
      "Entities:                           2   P-value                           0.0000\n",
      "Avg Obs:                    4.965e+04   Distribution:                 F(9,99287)\n",
      "Min Obs:                       1919.0                                           \n",
      "Max Obs:                    9.739e+04   F-statistic (robust):             2199.8\n",
      "                                        P-value                           0.0000\n",
      "Time periods:                      11   Distribution:                 F(9,99287)\n",
      "Avg Obs:                       9028.0                                           \n",
      "Min Obs:                       2135.0                                           \n",
      "Max Obs:                    1.988e+04                                           \n",
      "                                                                                \n",
      "                                               Parameter Estimates                                                \n",
      "==================================================================================================================\n",
      "                                                Parameter  Std. Err.     T-stat    P-value    Lower CI    Upper CI\n",
      "------------------------------------------------------------------------------------------------------------------\n",
      "Fully Interdisciplinary                            2.2353     0.2164     10.329     0.0000      1.8112      2.6595\n",
      "Interdisciplinary (Natural Sciences/Medicine)      1.3413     0.1546     8.6732     0.0000      1.0382      1.6444\n",
      "Interdisciplinary (Social Sciences/Humanities)     2.1753     0.1916     11.354     0.0000      1.7998      2.5508\n",
      "team_size                                          0.1424     0.0307     4.6369     0.0000      0.0822      0.2026\n",
      "article_length                                    -0.0003   1.98e-05    -13.685     0.0000     -0.0003     -0.0002\n",
      "cs.cl                                             -2.7096     0.1812    -14.957     0.0000     -3.0647     -2.3545\n",
      "cs.cv                                             -3.8735     0.1506    -25.715     0.0000     -4.1687     -3.5782\n",
      "cs.ai                                              1.5230     0.1246     12.223     0.0000      1.2788      1.7673\n",
      "cs.cy                                              31.635     0.3002     105.37     0.0000      31.047      32.224\n",
      "==================================================================================================================\n",
      "\n",
      "F-test for Poolability: 131.85\n",
      "P-value: 0.0000\n",
      "Distribution: F(11,99287)\n",
      "\n",
      "Included effects: Entity, Time\n",
      "\n",
      "PanelOLS Results with Clustered Standard Errors (by event_type):\n",
      "                          PanelOLS Estimation Summary                           \n",
      "================================================================================\n",
      "Dep. Variable:       percent_societal   R-squared:                        0.3674\n",
      "Estimator:                   PanelOLS   R-squared (Between):              0.0493\n",
      "No. Observations:               99308   R-squared (Within):               0.3609\n",
      "Date:                Thu, Jul 03 2025   R-squared (Overall):              0.3232\n",
      "Time:                        16:47:43   Log-likelihood                -3.954e+05\n",
      "Cov. Estimator:             Clustered                                           \n",
      "                                        F-statistic:                      6406.4\n",
      "Entities:                           2   P-value                           0.0000\n",
      "Avg Obs:                    4.965e+04   Distribution:                 F(9,99287)\n",
      "Min Obs:                       1919.0                                           \n",
      "Max Obs:                    9.739e+04   F-statistic (robust):          2.906e+19\n",
      "                                        P-value                           0.0000\n",
      "Time periods:                      11   Distribution:                 F(9,99287)\n",
      "Avg Obs:                       9028.0                                           \n",
      "Min Obs:                       2135.0                                           \n",
      "Max Obs:                    1.988e+04                                           \n",
      "                                                                                \n",
      "                                               Parameter Estimates                                                \n",
      "==================================================================================================================\n",
      "                                                Parameter  Std. Err.     T-stat    P-value    Lower CI    Upper CI\n",
      "------------------------------------------------------------------------------------------------------------------\n",
      "Fully Interdisciplinary                            2.2353     0.0231     96.926     0.0000      2.1901      2.2805\n",
      "Interdisciplinary (Natural Sciences/Medicine)      1.3413     0.0799     16.794     0.0000      1.1847      1.4978\n",
      "Interdisciplinary (Social Sciences/Humanities)     2.1753     0.0643     33.824     0.0000      2.0493      2.3014\n",
      "team_size                                          0.1424     0.0411     3.4643     0.0005      0.0618      0.2230\n",
      "article_length                                    -0.0003  5.849e-07    -463.25     0.0000     -0.0003     -0.0003\n",
      "cs.cl                                             -2.7096     0.0433    -62.626     0.0000     -2.7944     -2.6248\n",
      "cs.cv                                             -3.8735     0.0837    -46.271     0.0000     -4.0375     -3.7094\n",
      "cs.ai                                              1.5230     0.0341     44.613     0.0000      1.4561      1.5900\n",
      "cs.cy                                              31.635     0.0711     445.17     0.0000      31.496      31.775\n",
      "==================================================================================================================\n",
      "\n",
      "F-test for Poolability: 131.85\n",
      "P-value: 0.0000\n",
      "Distribution: F(11,99287)\n",
      "\n",
      "Included effects: Entity, Time\n"
     ]
    }
   ],
   "source": [
    "#### 80% threshold\n",
    "\n",
    "# Define dependent and independent variables\n",
    "Y = data_panel['percent_societal']\n",
    "X = pd.get_dummies(data_panel['interdisciplinary_category_80'], drop_first=True)  # Use CS-only as baseline\n",
    "X['team_size'] = data_panel['team_size']\n",
    "X['article_length'] = data_panel['article_length']\n",
    "\n",
    "# Add AI subfield controls if present\n",
    "for category in selected_categories:\n",
    "    if category in data_panel.columns:\n",
    "        X[category] = data_panel[category]\n",
    "\n",
    "# ---------------------------\n",
    "# Estimate Models\n",
    "# ---------------------------\n",
    "\n",
    "# Define the base model with fixed effects\n",
    "panel_model = PanelOLS(Y, X, entity_effects=True, time_effects=True)\n",
    "\n",
    "# Model 1: Robust standard errors (default recommended)\n",
    "results_robust = panel_model.fit(cov_type='robust')\n",
    "print(\"\\nPanelOLS Results with Robust Standard Errors for 80 Threshold:\")\n",
    "print(results_robust.summary)\n",
    "\n",
    "# Model 2: Clustered standard errors (not reliable with only 2 clusters)\n",
    "clusters = pd.Series(data_panel.index.get_level_values('event_type'), index=data_panel.index)\n",
    "results_clustered = panel_model.fit(cov_type='clustered', clusters=clusters)\n",
    "print(\"\\nPanelOLS Results with Clustered Standard Errors (by event_type):\")\n",
    "print(results_clustered.summary)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1f25d4b1",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "societal_influence_env",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.21"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}
